The connections between different layers of the cat's primary visual cortex are thought to play a critical role in shaping response properties, yet nothing is actually known about their functions. I am currently developing a method for studying one major interlaminar pathway, from layer 6 to 4. The curvature and thickness of layer 6 in area 17 make it possible to lesion an extensive portion of it without damange to more superficial neurons, or to fibers of passage. This can be achieved by injection of a cytoxic agent specific for neuronal cell bodies, kainic acid. Two major aims of this proposal are to carry out control experiments relevant to this approach, and to study layer 4 response properties following destruction of input from layer 6. Another important aspect of visual cortical organization about which little is known concerns the input to different areas of visual cortex. Although a description of the response properties of such inputs is critical for understanding the information processing carried out in any area of visual cortex, anly the geniculocortical pathway has been described in any detail. This lack stems from the inadequacies of present techniques dependent on antidromic activation of afferent cells; such experiments often yield only meagre data in return for a substantial investment in time and effort. Using a new approach, I have found that large injections of kainic acid into area 17 destroy all neurons within a restricted region, making it possible to record directly from geniculate and intracortical afferents terminating there; these remain unaffected. I propose to use this method to characterized the inputs to different areas of visual cortex.